Griffiths et al.: Fish recolonization in temperate Australian rockpools 



641 



Bathygobius cocosensis 



Enneapterygius rulopileus 



Spring 



^-j*— i*-! ^r 1 ^ 



Long-term 



Site 



I 



I 



i 



I 



Girella elevata 



Spring 



I 



Hsf^Ft fj 



ftA 



I 



ff^M 



\ 



5 

 4 

 3 



2- 

 1 - 



Long-term 



[fitt 



I 



If 



MB 



I 



"Y~ 



- T - 

 BB 



[~J Initial 



H 1 week 



!\] 1 month 



H 3 months 



7J Initial 

 PI 6 months 

 ffl 12 months 



Figure 4 



Mean l±SEl numbers of fish (/m 3 ) for the six most abundant species caught from rockpools at four sites 

 (MB=Maloney's Bay, TC=The Chair, GL = Gravel Loader, BB = Beaky Bay) during short-term (spring and autumn) 

 (four sampling events) and long-term (three sampling events) recolonization studies undertaken between 7 

 September 1999 and 22 September 2000. 



Table 5 



Results of repeated-measures ANOVAs for significant differences in numbers of individuals (/m 3 ) representing the six most abun- 

 dant species caught at Bass Point during the short-term recolonization studies among sampling intervals (within-subjects factor) 

 and sites and seasons lamong-subjects factors). Data were log 1(1 (.v+l) transformed before analysis to remove heteroscedasticity in 

 the data. Mauchly's criterion for sphericity of variances was violated (P<0.001) for species denoted by ''; therefore analysis was 

 performed using Greenhouse-Geisser-adjusted degrees of freedom. Greenhouse-Geisser degrees of freedom used for within-sub- 

 jects factors where Mauchly's criterion for sphericity of variances was violated lP<0.001): Time (Ti= 1.55; SexT = 1.55; SxT = 

 4.64; SexSxT = 4.64; Residual = 34.14. Mean squares and significance levels are shown and significant results are given in bold. 

 Degrees of freedom are shown in parentheses. * = P<0.05; ** = P<0.01; *** = P<0.001. 



Among-subjects factors 



Within-subjects factors 



Discussion 



This study has shown that fish assemblages can quickly 

 return to preperturbation levels after significant distur- 

 bance. This resilience appears to be driven mainly by 

 postsettlement movements of fishes, although recruit- 



ment may periodically play a significant role in popula- 

 tion replenishment. Not only does this provide an insight 

 into the ecology of rockpool fish assemblages, but this 

 information may also provide a basis for future sampling 

 protocols where the confounding effects of sampling may 

 be minimized. 



